Community-level determinants of infant mortality in Mexico

Data from pregnancy histories collected by the 1976-77 Mexican Fertility Survey show wide variations in infant mortality among a sample of 125 communities. For example, in regard to community population, localities of 2500 or less averaged 104/1000 live births which was almost 2 times the rate of 56/1000 experienced by cities with 100,000 or more people. While the rates for Mexico's 3 largest cities (Guadalajara, Monterrey, and Mexico City) were slightly higher, this trend was due to the large slum populations. Overall, there were large and statistically significant infant mortality differentials by community population, proportion of the labor force in agriculture, mean wage, access to the nearest urban center, communication means, proportion of households with electricity, piped water and sewers, and distance to the closest 2ndary and preparatory school and university. Considering health care variables, there was an expected relationship between distance to the nearest health care facility and mortality levels. These community-level factors, which have been litte used in previous studies, are highly correlated and do not appear to affect infant survival independently of population size. Community size serves as a summary measure of a locality's overall level of development and comparative risk of early death for its children.